Workhorses

May 19, 2020
                                                      Workhorses

            There was a question which arose in the "Hey, Bill" section on May 15th which I could not resist the urge to research.  The question had to do with whether Gene Mauch was noted for working his pitchers to their limits; actually, not to do with whether he was noted for this, but with whether he did this in fact.   It started with Art Mahaffey.  Anyway, I said that I didn’t think so; I thought that if you made up a "Index of Workload Scores" for managers, he would probably be in the middle of the scale.  So then I thought, well, that sounds interesting; why don’t I just do that?

            It turns out (spoiler alert) that Mauch (a) is in the middle of the scale, but (b) is clearly above average, meaning that he worked his pitchers a little harder than the average manager—but mostly relief pitchers, not the starters, and he is sort of in the middle of the scale, sort of toward the high end. 

            What we are interested in here is whether a manager works his pitchers hard compared to the norms of the era.   Everybody knows; we don’t have any Robin Robertses around today.  That’s not the issue; the issue is "compared to the norms of the era."  To measure that, I created a 5-element index.  The five elements are:

1)     R1 (Relief 1)  Games pitched. 

2)     R2 (Relief 2)  Relief Work Score; I’ll explain in a minute.

3)     S1 (Starter 1)  Innings pitched

4)     S2 (Starter 2)  Batters Faced

5)     S3 (Starter 3)  Batters Faced in Excess of fifteen per game.

 

The Relief Work Score is "Two times innings pitched, plus games, among pitchers whose batters faced per game appearance is no higher than 15.00." 

The "Batters Faced in excess of 15 per game" (S3) is like this.  Suppose that there are two pitchers who each face 900 batters in a season.   One pitcher, however, has faced 900 batters in 35 starts, and the other has faced 900 batters in 25 starts.  Who has the higher workload?

The key is that we are focusing with this measurement on long outings; that is, sending the pitcher repeatedly into the late innings.   If a pitcher faces 900 batters in 35 outings, that is 25.7 batters per outing.  If he faces 900 batters in 25 outings, that is 36 batters per outing, which is quite different.  The 35-outing pitcher has exceeded 15 batters per game by 375 batters.  The 25-outing pitcher has exceeded 15 batters per game by 525 batters, so he has a higher R3 score. 

With the exception of these two categories is just simple, straightforward stuff, the question being "how do you make an index out of this?"

I started. . . .

First of all, I should explain; I put the back border of the study in 1946.    I started with the top 25 pitchers in the majors in each year.   The pitcher who had the highest total in each of the five categories got 25 points; second-highest, 24 points, third-highest, 23 points, etc.   Actually it isn’t that the pitcher got the points as much as his manager did; we’re studying managers here. 

With 16 teams, that makes 325 points, or 20.3 points per team.  (25+24+23. . .. +3+2+1=325).  I kept the ratio about 20 points per team, so that managers over time could be compared to one another.   When the number of teams went to 18, points went to 26-25-24, etc.   When the number of teams went to 20 (1962), the number of points went to 28-27-26-25, etc.  When the number of teams went to 24 (1969), the number of points went to 30-29-28-27-26, etc.  When the number of teams went to 26 (1977) the number of points went to 32-31-30-29-28, etc.   When the number of teams went to 28 (1993) the number of points went to 33-32-31-30-29-28, etc.  When the number of teams went to 30 (1998) the number of points went to 34-33-32, etc.   The average number of points per team stays close to 20.00, so an average manager is going to score about 20.   If the manager scores higher than that, then he works his top pitchers harder than the average manager of the time.  If his score is lower than that, then he uses his pitchers in a cautious manner.   Make sense?

There are 20 points per category per team, and there are five categories, so the average manager will have very close to 100 points per season in the index of how hard he works his pitchers.  The exact average is 100.13.   This was not the intent of the study; I started with 20 points and one category, but then I added a second, a third. … .better get the relievers.  I wound up with five categories, thus 100 points per manager per season, but that was just an accident.

Since we are depending on rank-order points here, there has to a "second sort" for each category; otherwise you get a lot of ties.  The second sorts are:

R1 (Games)  Second sort is innings pitched.

R2 (Relief Work Score) Second sort is Innings Pitched.

S1 (Innings Pitched)  Second sort is fewer games; that is, a pitcher who pitches 200 innings in 30 games ranks ahead of a pitcher who pitches 200 innings in 31 games.

S2 (Batters Faced)  Second sort is fewer games.

S3 (BFP in excess of 15 per game)  Second sort is BFP, Batters Facing Pitcher. 

 

Since 1946, a total of 5,546 pitcher/seasons have figured in the study, meaning that the pitcher earned at least one point in one of these categories.  Those 5,546 seasons have been by 1,869 different pitchers, and, for whatever this is worth, a total of 180,827 "high workload points" have been awarded over the 74 years of the study.  Of course, the pitcher who finished first in one category often also finishes first in a second category.   I’m just trying to look at the issue in different ways; it’s my measurement philosophy.   The concept of "size" means height, weight, and body mass index.   You don’t measure one of those things and ignore the others.

 

Before we come to the real purpose of our study, we come to the question of "Who are the hardest working pitchers within the study?"  I emphasize that this is NOT what we are studying here; if I was studying pitchers I would have designed the study in a slightly different way.  I am studying managers, not pitchers.  However, we’re here; I have this data, and it implicates an interesting if not-immediately-relevant question.  I might as well share with you what I have, even though I will warn you in advance that the system was not designed to answer this question, and the answers that it gives to this question are generally sensible, but not completely persuasive. 

Before we get to the question of the hardest-working pitchers, we come to the yet-less-interesting and yet-more-irrelevant question of the highest scores in each category for a SEASON.   Those answers are straightforward, and many of you could guess most of them, but. . . .for whatever it is worth.

R1, the most games pitched in a season, is, of course, Mike Marshall in 1974.  Marshall is followed by Kent Tekulve (1979), Solomon Torres (2006), Mike Marshall (1973), Pedro Feliciano (2010), Kent Tekulve (1978), Kent Tekulve (1987) and Mike Marshall (1979). 

R2, the highest reliever workload score in a season is, of course, Mike Marshall in 1974.  He is followed on that list by Mike Marshall, 1973, Wilbur Wood, 1968, Eddie Fisher, 1965; Bill Campbell, 1976; Wayne Granger, 1969; Mike Marshall, 1979; and Dick Radatz, 1964.  Then you get, interestingly enough, Ferguson Jenkins, 1966; Fergie that year made 12 starts but 49 relief appearances, 61 games, 184 innings, which you have to admit is a pretty good workload.  Then we have Kent Tekulve, 1979, and Kent Tekulve, 1978. 

S1, Innings pitched, the highest number is actually a tie between Wilbur Wood, 1972, and Mickey Lolich, 1971, each of whom pitched 376.1 innings.   After them we have Bob Feller, 1946; Wilbur Wood, 1973; Robin Roberts, 1953; Steve Carlton, 1972; Gaylord Perry, 1973; Gaylord Perry, 1972; Phil Niekro, 1979, and Robin Roberts, 1954. 

S2, Batters Faced, the highest number is Mickey Lolich, 1971, 1,538, followed by Wilbur Wood, 1973.   Then you have the same list of pitchers that we had before in slightly different order, with the exception that Nolan Ryan, 1974, pushes onto the list because he walked so many people that it took him extra batters to get the same number of innings. 

S3 Batters Faced in Excess of 15 per game, the number one man again is Mickey Lolich, and this time by a much wider margin.   Lolich is at 863; no one else is over 796.    Following Lolich on that list are Wilbur Wood, 1973; Gaylord Perry, 1973; Bob Feller, 1946; Phil Niekro, 1979; Phil Niekro, 1977: Nolan Ryan, 1974; Wilbur Wood, 1972; Robin Roberts, 1953; Bill Singer-Songwriter, 1973; Gaylord Perry, 1969; Bob Gibson, 1969; and Nolan Ryan, 1973. 

 

We come now to the question of the hardest-working pitchers IN A CAREER, by this method (which I have acknowledged, and will acknowledge again, is not perfectly suited to answer this question, because this is not exactly what we are studying.)    What we really mean by a "workhorse pitcher", though, is not so much a pitcher who pitches a lot of games or innings in one season, but rather, a pitcher who can do it year after year.   Modern pitchers like Jon Lester, Max Scherzer, Justin Verlander and Zack Greinke are on the hardest-working pitchers list year after year, as were the other great pitchers in their time—Spahn and Clemens and Jim Kaat, etc.  

OK, these are the top 50 "Workhorse Pitchers" since 1946, by this (imperfect) method, acknowledging in addition that Bob Feller would make the list despite missing the years 1942-1945 if we had not started the study in the middle of his career:

 

Rank

First Name

Last

1

Warren

Spahn

2

Greg

Maddux

3

Randy

Johnson

4

Steve

Carlton

5

Roger

Clemens

6

Justin

Verlander

7

Jack

Morris

8

Gaylord

Perry

9

Bert

Blyleven

10

Livan

Hernandez

11

Robin

Roberts

12

Tom

Glavine

13

Ferguson

Jenkins

14

Phil

Niekro

15

James

Shields

16

Tom

Seaver

17

Roy

Halladay

18

Mark

Buehrle

19

Don

Drysdale

20

John

Smoltz

21

Mark

Langston

22

Jim

Palmer

23

Frank

Viola

24

CC

Sabathia

25

Early

Wynn

26

Nolan

Ryan

27

Bob

Gibson

28

Chuck

Finley

29

Fernando

Valenzuela

30

Jim

Bunning

31

Wilbur

Wood

32

Charlie

Hough

33

Juan

Marichal

34

Bob

Lemon

35

Kent

Tekulve

36

Curt

Schilling

37

Felix

Hernandez

38

Dennis

Martinez

39

Kevin

Brown

40

Tim

Hudson

41

Larry

Jackson

42

Rollie

Fingers

43

Kenny

Rogers

44

Bob

Friend

45

Mike

Mussina

46

Zack

Greinke

47

Dan

Haren

48

Cole

Hamels

49

Mickey

Lolich

50

Dave

Stieb

 

Of the top 10 workhorse pitchers since 1946, all are currently in the Hall of Fame except Roger Clemens and Livan Hernandez.  Of the second 10, all are in the Hall of Fame except James Shields, Mark Buerhle, and Roy Halladay, who has been elected but not yet officially inducted.   Of the third 10, five are in the Hall of Fame and five are not; of the fourth 10 and the fifth 10, only two each are in the Hall of Fame. 

It is surprising that Nolan Ryan is not higher on the list, no?   I’m not saying that this list is right and another one would be wrong.   Ryan doesn’t rank higher because, while his career is fantastic for its length, the number of seasons in which he was among the major league leaders in workload carried is less than you might guess.  Of course, when I hear the term "Workhorse Pitcher", I think more of Nolan Ryan than of James Shields, but that’s the reason you do research, just to see what you find.  Ryan has seven seasons in which he scores at 60 or above by this method.   Warren Spahn has thirteen.

Most of it is that "standards of the time".  Big Game James Shields was a workhorse pitcher by the standards of the last decade, which require that he pitches 200 innings to be a workhorse.  Nolan Ryan was a workhorse pitcher by the standards of the 1970s, when many pitchers pitched 300.  Different thing.  It’s necessary to reference the standards of the time while studying managers.   If I was studying pitchers directly, I’d probably reference the standards of the time in a different way.

 

Now we move on toward the managers, which is the main purpose of our study.   Let me give you my usual litany of caveats, data points and caution flags (no, I didn’t watch the damned car race on Sunday; I don’t watch NASCAR.)   Anyway, three of those:

1)     First, we have to attribute each pitcher to a team, even if he pitched for two teams in the season or three teams in the season, and then we have to attribute each team to a manager, even if the team changed generals in the middle of the campaign.   Both of these attribution issues create possible misreadings of the data. 

Of course, 94 to 95% of pitchers who have high workloads do not change teams in the middle of the season, and, of the other 5 to 6%, the great majority do so either very early in the season or very late, but the other 1% of the cases are problematic.  I remember that there was one pitcher who pitched 33 games and 50 innings with one team and 33 games and 49.2 innings with the other team, but the study requires you to put him with one team or the other.   Also, there was one team that split a 154-game season 77 games with one manager, 77 with the other, so. .  ..those are things you wish you didn’t have to deal with, but you do, and they can create misleading data. 

 

2)     There are 1,806 seasons of team data in the study, thus 1,806 manager/seasons once each season is attributed to one manager.  Those seasons are by 314 different managers, or 100 pies worth.  Of those 314 managers, 59 managed 10 seasons or more within our data.  Those 59 managers will form our core group, and I’ll give you the data for all of them. 

3)     You might suspect that good teams tend to have higher workload scores for their top pitchers than bad teams do, and that MAY BE true at some level.  Take Walter Alston.  Alston has a high yearly average pitcher Workload Score, 137—but Alston had Don Drysdale and Sandy Koufax.   If you had Sandy Koufax and Don Drysdale, you’d let them pitch a lot, too.  But the data shows that, while there is probably a bias in that direction, it isn’t a major problem.   More later.

 

OK, the data.  These are the highest scoring managers of all time:

Rank

Manager

Years

Relief Points

Starting Points

Total

Relief Average

Starter Average

Average

1

Bobby Cox

29

1622

3594

5216

56

124

180

2

Billy Martin

15

649

1890

2539

43

126

169

 

Let’s stop right there.   The argument that many of you are making in your head is that Bobby Cox "only" ranks higher than Billy Martin because Cox HAD all of those wonderful starting pitchers—Maddux and SmoltzMouth and Glavine.   When you have those pitchers, you pitch them.   A lot. 

That’s not entirely untrue, but it’s not entirely true, either.   Cox with the Blue Jays had scores of 220, 204 and 203 (1982-84-85) when he didn’t have pitchers anything LIKE Maddux, Smaltzie and Glavine.   In 1982 one of his starting pitchers led the league in starts (40), while another one led the league in innings pitched and batters faced.  One of those was a great pitcher, yes (Dave Stieb), but his Smoltz and Glavine were Jim Clancy and Luis Leal.   Clancy—128-140 in this career with a 4.10 ERA—pitched 267 innings for Cox in 1982, 223 in 1983 and 220 in 1984.  Luis Leal—51-58 in his career with a 4.14 ERA—pitched 250 innings for Cox in 1982, 217 in 1983, and 220 in 1984.    Doyle Alexander—averaging 138 innings per season from 1981 to 1983—pitched 262 innings in 1984 and 261 in 1985, when managed by Cox.  He was third in the league in innings pitched one year, fourth the other. 

Second on the list is the guy that most of you would have expected to be first:  Billy Martin.   I’m not saying that Bobby scoring higher than Billy is absolutely right and another answer would be absolutely wrong, but here’s one way to look at it. 

First, if a new system, a new index. .. if that index ranks the most obvious candidate to be first on the list as SECOND on the list, that’s more of a success than it is a failure. 

            Second, we are not talking here about whether any manager "abused" his best pitchers; that is a totally separate issue.  We’re measuring the extent to which each and every manager relied heavily on his best pitchers, without any judgment as to whether this was right or wrong.

Billy Martin became famous as the guy who worked his pitchers really hard essentially because of two cases in which he over-extended his stars to an almost comic degree—Mickey Lolich in 1971-72, when Lolich pitched a little over 700 innings in two seasons, and Oakland in 1980-1981, when Martin pushed a staff of four young starters to their limits, starting a controversy about whether this was appropriate.  (It wasn’t.) 

But almost half of Martin’s career was with the Yankees, in between many hirings and firings.  With the Yankees, where Martin had more options to work with, Martin’s pitcher usage was high end, but not absurdly high end—not more high-end than Cox. 

In 1980, Martin pushed the 28-year-old Rick Langford to throw 28 complete games and 290 innings, which was over the top (as was Lolich’ performance nine years earlier.)  That’s famous.   But two years later, Cox pushed the 24-year-old Dave Stieb to pitch 19 complete games and 288 innings.  That’s not a whole lot different.  It’s not famous, not infamous, but it’s not a whole lot different. 

But actually, dealing just with STARTING pitchers, Martin’s score is higher than Cox’s—126 to 124.   The average on that part of the scale is 60; these guys are over 120—meaning that they have more than twice the normal number of pitchers who are at the top of the league list in terms of innings, batters faced, and batters faced in excess of 15 per game.  The reason that Cox is number one on the list and Cox is number two is that he used his bullpen leaders more heavily than Martin did; not the starters, the bullpen.   Cox’s average score for relievers was 56; Martin’s was 43.

Anyway, returning to the top 10 list:

Rank

Manager

Years

Relief Points

Starting Points

Total

Relief Average

Starter Average

Average

1

Bobby Cox

29

1622

3594

5216

56

124

180

2

Billy Martin

15

649

1890

2539

43

126

169

3

Jim Tracy

11

969

717

1686

88

65

153

4

Al Lopez

15

737

1470

2207

49

98

147

5

Chuck Tanner

17

1128

1211

2339

66

71

138

6

Walter Alston

23

1111

2046

3157

48

89

137

7

Al Dark

13

820

956

1776

63

74

137

8

Leo Durocher

17

854

1360

2214

50

80

130

9

Dusty Baker

22

1317

1508

2825

60

69

128

10

Earl Weaver

17

149

2021

2170

9

119

128

 

            Jim Tracy has used his best relievers harder than any other manager in the study, compared to the norms of the era.  In his first year as a manager, with the Dodgers in 2001, Jeff Shaw pitched in 77 games and Matt Herges in 75.   In his second year he brought Paul Quantrill into the game 86 times, Eric Gagne 77 times, Giovanni Carrara 63 times, and Jesse Orosco 56 times.  His third year, he upped the ante to 89 games for Quantrill, Paul Martin 80 times, Eric Gagne another 77, AND Guillermo Mota 76.   And Paul Shuey another 62 just for good measure, not that that figures in the scoring.  

            This is a quite unusual workload for a group of relievers.  It’s historic.   It’s the Billy Martin workload of relief pitchers.   His last year as a manager, 2012 with the Rockies, he called for Matt Belisle 80 times, Rex Brothers 75, Matt Reynolds 71, Josh Roenicke 63, Rafael Betancourt 60, and Adam Ottavino 53.   But you know; if you had Matt Belisle on your staff, you’d pitch him a lot, too.  I guess. 

            Al Lopez’ great success as a manager was driven largely by his unusual ability to keep his starting pitchers healthy and in the rotation.  In 1951, his first year as a major league manager, his four starting pitchers finished first, second, fourth and fifth in the league in innings pitched.   In 1952, his second year as a manager, his pitchers finished first, second and third in the league in innings pitched (Bob Lemon, 310, Mike Garcia, 292, and Early Wynn, 286.)  In 1953 his pitchers were first and second in the league in innings; in 1954 first, third and fourth. 

            While Lopez did moderate his heavy usage tendencies later in his career, he was always relatively near the top of the heavy-usage spectrum.  But also, late in his career, he made very heavy use of his top relievers.   Eddie Fisher in 1965 pitched 82 games, 165 innings—as a reliever.  Hoyt Wilhelm pitched 66 games, 144 innings for the same team.   Two relievers pitched 309 innings.   While Lopez CAREER bullpen score (49) is not exceptionally high, he did substitute heavy usage of his top relievers for super-heavy use of his starters.

            Chuck Tanner used a three-man starting rotation in portions of the 1971, 1972 and 1973 seasons.   In 1972—a strike-shortened season, by about a week—he gave 130 starts to three pitchers (Wilbur Wood, Stan Bahnsen and Tom Bradley) who pitched almost 900 innings among them.

            Walter Alston is perhaps a surprise entry on the list, but think about it.  Sandy Koufax’ early retirement came after he had pitched 311, 336 and 317 innings in three of the previous four seasons.   In 1962 Don Drysdale (pitching for Alston) led the league in starts (41) and innings (314). In 1963 Drysdale tied for the league lead in starts, again, while Drysdale and Koufax were second and third in the league in innings (315 and 311).  In 1964 Drysdale led the league again in starts (40) and innings (321).   In 1965 Koufax and Drysdale were 1-2 in the league in starts and innings pitched.  In 1966 Koufax tied for the league lead in starts (41), while Drysdale was one behind at 40; Koufax led the league in innings pitched, and Drysdale was among the league leaders.

            Again, this is not a judgmental article; it is a fact-based systematic analysis—but it is a fact that Drysdale and Koufax both retired early.  In 1967, with Koufax retired, Claude Osteen and Drysdale were 4th and 5th in the league in innings.

            And, like Lopez, Alston transitioned to exceptionally heavy use of his top relievers later in his career.  It was Alston who used Mike Marshall 106 games, 208 innings in relief in 1974.   In Alston’s first year as a manager, 1954, one of his relievers (Jim Hughes) led the league in games pitched.  Alston relievers led the league in game appearances in 1954, 1955, 1962, 1963, 1964, 1967 and 1974—five different relievers, and Ron Perranoski several times.

            Al Dark and Dusty Baker made only slightly heavier-than-average use of their top starting pitchers, but had high usage numbers for their top relievers.  Leo Durocher would rank higher on this list than he does if the earlier portion of his managerial career (1939-1945) was within the scope of the study.

            Earl Weaver—who had the four 20-game winners in 1971—had the highest use scores for his STARTING pitchers of any manager in the study except Martin and Cox, but had the LOWEST heavy-use score of any manager in the study for his relievers.  While he often did have effective relievers, they generally didn’t pitch a lot of innings, and were never the stars of the team.   Pete RIchert (in Weaver’s bullpen) had ERAs of 2.20 in 1969 and 1.98 in 1970, but pitched only 44 times in 1969, 51 times in 1970, with less than 60 innings each year.  Eddie Watt, for Weaver, had ERAs of 2.27 (1968), 1.65 (1969), 1.82 (1971), and 2.17 (1972)—but never pitched as many 60 games in a season or recorded more than 16 saves.  Weaver preferred complete games to saves.  His average relief work score was 9—the lowest in the study.  Eric Wedge was second-lowest, at 14.  (Doug Rader, Ozzie Guillen and George Bamberger had lower scores, but none of them met the ten-seasons standard.) 

 

            Now turning our attention to the BOTTOM of the list; these are the 10 managers who were LEAST inclined to rely heavily on their top pitchers:

Rank

Manager

Years

Relief Points

Starting Points

Total

Relief Average

Starter Average

Average

1

Casey Stengel

16

285

637

922

18

40

58

2

Buck Showalter

20

385

902

1287

19

45

64

3

Jim Fregosi

14

484

541

1025

35

39

73

4

Fred Hutchinson

11

206

620

826

19

56

75

5

Paul Richards

12

361

552

913

30

46

76

6

Mike Hargrove

15

453

708

1161

30

47

77

7

Eric Wedge

10

143

648

791

14

65

79

8

Ralph Houk

20

416

1183

1599

21

59

80

9

Birdie Tebbetts

10

473

333

806

47

33

81

10

Chuck Dressen

10

267

542

809

27

54

81

 

            Casey Stengel relied less heavily on his top pitchers than any other manager in the study—clearly less, and much less.  Whitey Ford hated it.  Whitey wanted to just step into the rotation and pitch every fourth day, like the other top starters of the era did.    Casey wanted to save him, spot him against better teams, and make sure that he was strong in September and October. 

            But it wasn’t just Whitey; Casey generally never used any pitcher a lot of innings or a lot of games.  He did some early in his career; in 1949, his first year in New York, Joe Page led the league in games pitched.  Vic Raschi had fairly high numbers of starts and innings.  But Ryne Duren, a hard-throwing sensation with strikeout numbers in 1950s that would still be considered high TODAY, pitched only 44 games in 1958, 41 in 1959, despite being tremendously effective. 

            And here we get to the other side of the argument, the argument that managers like Cox and Alston only had high-usage totals because they had great pitchers to work with.  Casey was as successful with the Yankees as any manager has ever been, but he didn’t do it that way. 

            I use this example.  Let’s go back to Red Ruffing.  Red Ruffing pitched 1925 to 1929 for the worst team in the American League, the Red Sox.  For the Red Sox, his winning percentages were WORSE than his team’s.  From 1925-1929 Ruffing went 9-18, 6-15, 5-13, 10-25 and 9-22, leading the league in losses in 1928.  But when Ruffing was traded to the Yankees, all of a sudden his winning percentage became BETTER than his team’s!   How is that possible?   That puzzled me for years—that a pitcher who was worse than his team on the worst team in the league could be better than his team on the best team in the league.  

            But I finally realized what it was.   With the Red Sox, Ruffing was over-worked.  He had the best arm in the American League, and the Red Sox had nobody else worth a dime to put on the mound, so they ran Ruffing out there again, and again, and again, and they let him pitch until the game was irretrievably lost, because they had nobody in the bullpen.  With the Yankees, Ruffing shared the burden with Lefty Gomez and Johnny Allen and other very fine pitchers.  There was no need to overuse him.

            Casey Stengel was successful with the Yankees because his style was to rotate players in and out, picking their spots for the weaker players and limiting the load for the better ones (other than Yogi.)   It didn’t work with a weak team, but it worked with the Yankees. 

            Billy Martin and Casey Stengel had a special relationship.  Casey managed Billy with the Oakland Oaks in 1948, and brought him to the major leagues in 1950.  Billy—an entertaining speaker—had a million Casey Stengel stories, and seemed very fond of him.

            Often a manager adopts the style of his mentor—but this is a case in which the opposite is clearly true.   Well, for one thing, Leo Durocher was more Billy’s mentor than Casey was; Billy knew Leo through Cookie Lavagetto, who had played for Leo and was Billy’s teammate and running mate in Oakland.  But anyway, it is interesting to see Billy and Casey on opposite ends of this scale.  You might not expect to see that. 

            Bobby Cox was also a Yankee product; his only major league experience was playing for Ralph Houk in the mid-to-late sixties.   He’s on the other end of the scale, too.  

Buck Showalter.    Generally, every team has SOME pitcher on the high-usage lists every year, maybe at the bottom of the list, but somebody.  Among the 1,806 teams in the study, there are only 32 teams which had a score for the year of zero. 

Buck Showalter managed three of those teams—the Texas Rangers in 2005, and the Orioles in 2012 and 2015.    In 2012 the Orioles won 93 games.  Their top starting pitcher was Wei-Yin Chen, who was 12-11 with a 4.02 ERA in 193 innings.  He was the only pitcher on the team who made more than 20 starts. 

3. Jim Fregosi and 4. Fred Hutchinson.   No comments; no insight. 

5.  Paul Richards.  Paul Richards, for most of his career, had the reputation for being a wizard with starting pitchers.  I once did a study of 20-game winners by manager, and discovered to my absolute astonishment that Richards had never managed a 20-game winner, not even once.  (He managed Virgil Trucks, a 20-game winner in 1953, for most of the season, but Trucks already had 5 wins in the bank before the White Sox traded for him.)  But now I understand the reason for this remarkable split.   Richards just didn’t push his best starting pitchers to their limits. 

Richards damaged his legacy by trying to manage the White Sox in the late 70s, when the game had passed him by, and also, he was not a straightforward person; he was a manipulator—which most good managers are—and a deceiver, which most are not.  But for most of his career, he had a real ability to bring pitchers along, building their confidence and keeping them within limits where they could be effective. 

6-7  Mike Hargrove and Eric Wedge. 

8.  Ralph Houk.   This was a huge surprise to me, because I had always thought of Houk as a manager who worked his starting pitchers very hard.  I had numerous reasons for that—Ford and Bill Terry in ’61-’63, Mel Stottlemyre in Houk’s return to the Yankees, and Mark Fidrych in ’76; Houk let the 21-year-old phenom throw 24 complete games in 29 starts. 

The data shows, however, that Houk was not actually a high-use manager; he just seemed like a high-use manager compared to Casey Stengel.   Mel Stottlemyre pitched a lot of innings, but (a) so did other pitchers in that era, and (b) the only year Stottlemyre actually led the league in innings pitched, Houk wasn’t managing him.   

Chuck Dressen and Birdie Tebbetts, tied at 10 seasons within the frame of the study, and 81 points.  That’s just funny to me, because for years, I always confused Chuck Dressen and Birdie Tebbetts, could never remember which was which.  It wasn’t until I actually met Birdie, sitting next to him at a game about 1983, that I could remember which one was which.

The generalization that good teams have high scores may be 51% true, but it is 49% false.  Many fine managers with successful careers have had low scores on the Pitcher Workload Index.  Among managers with below-average scores are Lou Piniella, Bruce Bochy, Joe Torre, Whitey Herzog, Joe Girardi, Tony LaRussa, Tom Lasorda, Davey Johnson and Red Schoendienst, as well as Casey Stengel, Paul Richards, Fred Hutchinson and Ralph Houk.  There’s a lot of successful seasons in that group, a lot of World Championships.  I would guess there are significantly more World Series rings in the "below average" group than in the "above average group." 

A few words in closing about active managers and others in the 5-to-9 years group:

Mike Matheny, the Royals’ new manager, actually has a higher average score through his first six seasons than even Bobby Cox or Billy Martin.  He had some horses to ride in St. Louis; we’ll see what it is like in Kansas City.

Pete Rose, with only a 5-year managerial career, also had a higher average than Martin or Cox, in his case because of absurdly high numbers in the bullpen, averaging 103.   In 1986 he had two relievers with 70+ games AND 100+ innings.   In 1987 he had two relievers with 87 and 85 games AND 100+ innings.  His starting pitchers had well above-average workloads, as well.

A.    J. Hinch, seven years into his work, has an average score of 124.

Kevin Cash is on the low end, at 71.  (For purposes of this study, we don’t care whether an outing was actually a start or a relief appearance.  The "starter" and "reliever" classifications are based on Batters Faced per Appearance.)

These are the scores for all 59 managers within the study who were the primary manager for the season for at least ten teams:

Rank

Manager

Years

Relief Points

Starting Points

Total

Relief Average

Starter Average

Average

1

Bobby Cox

29

1622

3594

5216

56

124

180

2

Billy Martin

15

649

1890

2539

43

126

169

3

Jim Tracy

11

969

717

1686

88

65

153

4

Al Lopez

15

737

1470

2207

49

98

147

5

Chuck Tanner

17

1128

1211

2339

66

71

138

6

Walter Alston

23

1111

2046

3157

48

89

137

7

Al Dark

13

820

956

1776

63

74

137

8

Leo Durocher

17

854

1360

2214

50

80

130

9

Dusty Baker

22

1317

1508

2825

60

69

128

10

Earl Weaver

17

149

2021

2170

9

119

128

11

Frank Robinson

13

871

782

1653

67

60

127

12

Cito Gaston

12

345

1091

1436

29

91

120

13

Bill Virdon

13

376

1177

1553

29

91

119

14

Sparky Anderson

26

1436

1610

3046

55

62

117

15

Jimy Williams

11

890

398

1288

81

36

117

16

Phil Garner

13

499

1022

1521

38

79

117

17

Charlie Manuel

12

463

938

1401

39

78

117

18

Gene Mauch

24

1178

1589

2767

49

66

115

19

John McNamara

14

658

943

1601

47

67

114

20

Felipe Alou

13

956

520

1476

74

40

114

21

Joe Maddon

14

650

898

1548

46

64

111

22

Bob Melvin

15

630

1021

1651

42

68

110

23

Art Howe

14

938

583

1521

67

42

109

24

Terry Francona

19

719

1303

2022

38

69

106

25

Bill Rigney

16

972

713

1685

61

45

105

26

Dick Williams

16

554

1125

1679

35

70

105

27

Tom Kelly

15

516

1044

1560

34

70

104

28

Lou Boudreau

12

620

620

1240

52

52

103

29

Ned Yost

16

426

1220

1646

27

76

103

30

Jim Leyland

23

704

1635

2339

31

71

102

Rank

Manager

Years

Relief Points

Starting Points

Total

Relief Average

Starter Average

Average

31

Mike Scioscia

19

598

1320

1918

31

69

101

32

Danny Murtaugh

12

575

628

1203

48

52

100

33

Red Schoendienst

12

444

729

1173

37

61

98

34

Clint Hurdle

16

981

578

1559

61

36

97

35

Davey Johnson

16

502

1053

1555

31

66

97

36

Bobby Valentine

15

756

691

1447

50

46

96

37

Tom Lasorda

19

324

1500

1824

17

79

96

38

Tony LaRussa

32

883

2182

3065

28

68

96

39

Johnny Oates

10

194

741

935

19

74

94

40

Joe Girardi

11

270

756

1026

25

69

93

41

Don Zimmer

11

441

568

1009

40

52

92

42

Whitey Herzog

16

574

872

1446

36

55

90

43

Joe Torre

27

1585

814

2399

59

30

89

44

Terry Collins

13

667

482

1149

51

37

88

45

Bruce Bochy

25

493

1641

2134

20

66

85

46

Jack McKeon

13

650

453

1103

50

35

85

47

Lou Piniella

23

728

1220

1948

32

53

85

48

Ron Gardenhire

15

708

548

1256

47

37

84

49

Buddy Black

11

574

340

914

52

31

83

50

Chuck Dressen

10

267

542

809

27

54

81

51

Birdie Tebbetts

10

473

333

806

47

33

81

52

Ralph Houk

20

416

1183

1599

21

59

80

53

Eric Wedge

10

143

648

791

14

65

79

54

Mike Hargrove

15

453

708

1161

30

47

77

55

Paul Richards

12

361

552

913

30

46

76

56

Fred Hutchinson

11

206

620

826

19

56

75

57

Jim Fregosi

14

484

541

1025

35

39

73

58

Buck Showalter

20

385

902

1287

19

45

64

59

Casey Stengel

16

285

637

922

18

40

58

 

Thanks for reading. 

 
 

COMMENTS (15 Comments, most recent shown first)

LesLein
Thanks. Great article, but what about Art Mahaffey? If he was overused it was in 1962. As a 24 year old he had 1138 BFP. In 1964 Jim Bunning had 1145 BFP. Mahaffey was a youngster getting established on a team fighting to finish over .500. Bunning was a veteran and future HOFer for a team fighting for the pennant. It seems to me Mahaffey was overused. More details are needed to determine if he was abused.
7:22 PM May 20th
 
raincheck
bjames
“How would you come in for yourself in relief?”


It was a hilarious joke referring to his heavy usage.
6:36 PM May 20th
 
StatsGuru
So Zimmer only overworked his position players!
6:12 PM May 20th
 
pbspelly
I recall you speculating years ago, in one of your abstracts, that perhaps Martin overworked his pitchers based on his own experience of getting fired so often. He tended to view things short term because for him, it was always short term.
2:07 PM May 20th
 
Steven Goldleaf
I think raincheck was joking about coming in for himself in relief. The tip-off was "pitched every inning of every playoff game," which would be some kind of record.
1:28 PM May 20th
 
bjames
Following up on the previous post, about the standard deviation, I realize now how that could be done. You could assign a "workload score" to every pitcher in history; let's say that 40 games and 300 innings is a workload score of 80, 30 games and 200 innings in a workload score of 60, 20 games and 100 innings is a workload score of 40, 15 games and 30 innings is a workload score of 20, etc. Then you could establish the norms for every season in history, focus on the top 5 or top 8 pitchers on each team, and determine where each pitcher stood with regard to those norms, then derive values for managers from that. But that would be like 20 times more work than I have done. You'd have to tag every pitcher in history to a manager, rather than tagging 25 to 35 pitchers a year to a manager. Different level of effort, but you'd have the granularity of data which would enable you to study the year to year fluctuations for individual managers. I generally don't undertake studies of that magnitude, unless it's a major issue. Nobody does, actually, but maybe they will in the next generation.
12:44 PM May 20th
 
bjames
In regard to the standard deviation suggestion. . . the single-season data would not be stable enough for that to be meaningful. The system is designed to look for patterns over time. It doesn't get a sensitive measure for one year. I don't know how you would get a sensitive measure for one year; maybe you could, but certainly not by this approach.
12:05 PM May 20th
 
bjames
How would you come in for yourself in relief? But I always perceived Lasorda as a guy worked his starters hard, too. His score for starting pitchers IS actually quite high--79. Average is 60, so he's 30% above average for starting pitchers. But he didn't use his relievers hard.
11:55 AM May 20th
 
raincheck
I always perceived Lasorda to be a guy who used a pitcher a lot, based, for example, on how he he worked a young Fernando in his rookie year, how he rode Hershieser in 1988 (I may be wrong, I haven’t looked it up, but didn’t Orel pitch every inning of every playoff game, even coming in for himself in relief?)

It seemed like when he had faith in a guy that guy was going to work a lot. It is interesting to see him in the middle of the pack, I have to rethink my intuitive fan-based frustration with Tommy and how it seemed like the Dodgers ran through a lot of young pitchers. I don’t think I realized then that most young pitchers drop off or flame out and very few become a long term star.
11:02 AM May 20th
 
raincheck
I always perceived Lasorda to be a guy who used a pitcher a lot, based, for example, on how he he worked a young Fernando in his rookie year, how he rode Hershieser in 1988 (I may be wrong, I haven’t looked it up, but didn’t Orel pitch every inning of every playoff game, even coming in for himself in relief?)

It seemed like when he had faith in a guy that guy was going to work a lot. It is interesting to see him in the middle of the pack, I have to rethink my intuitive fan-based frustration with Tommy and how it seemed like the Dodgers ran through a lot of young pitchers. I don’t think I realized then that most young pitchers drop off or flame out and very few become a long term star.
11:02 AM May 20th
 
FrankD
Another great article. Very interesting that the top managers are sprinkled throughout the Work Horse spectrum. AS fro Stengel/Martin: maybe Billy heard a lot of pitchers complain about under usage while he was playing for Stengel and Billy thought that I'm going to ride my horses if I ever get in charge .....
8:00 AM May 20th
 
rtallia
I see that the HOF managers are sprinkled throughout the list, including the very top and very bottom, of course. Was wondering if you studied which managers had the biggest VARIANCE in their approach from year to year...assume it would be guys in the middle of the pack, who could then (potentially) be characterized as tailoring their approach given what they had to work with.
6:37 AM May 20th
 
AJD600
Would it be useful to do the same study on pitching coaches?
9:18 PM May 19th
 
W.T.Mons10
I'm not surprised Houk didn't score high in starter usage, but I thought he'd be higher in relief usage. He pitched Lindy McDaniel a lot a couple of years, and Luis Arroyo threw a lot in 1961, but I guess they were the exceptions. In 1961, when Ford started 41 games, no one else on the staff started even 28 games.
9:05 PM May 19th
 
SteveN
Surprised that their was no mention of Mark Eichhorn in the heavily used reliever category. Of course, he had only 2 years of severe usage. An old favorite.

I've seen, in person, a pretty high percentage of the heavily used starting pitchers. I guess I'm lucky.

Thanks
12:19 PM May 19th
 
 
©2020 Be Jolly, Inc. All Rights Reserved.|Web site design and development by Americaneagle.com|Terms & Conditions|Privacy Policy